Frequency Electronics is a precision time and frequency specialist that sold off a low-margin business at the end of fiscal 2026, took a one-time hit to its income statement, and let the order book nearly double in the same move. The argument for the stock is that the digestion is over. The backlog, now at roughly $111 million, is set to convert into reported growth through fiscal 2027.
The most important recent development is the registered offering completed on July 30, 2026. The company sold about 1.09 million new shares at $57.50 each. It raised roughly $62.5 million before underwriting discounts. Two Edenbrook Value funds sold 652,174 secondary shares in the same transaction. The mechanism matters: the company had barely any cash on hand after funding a deliberate revenue pull-forward and a restructuring, so the equity raise refills the war chest that the backlog ramp needs, and the secondary block trims the balance of a concentrated insider position. The stock price, in the mid-60s shortly after the print, already sits above the offering price, so the dilution was priced in before completion.
The central tension is customer concentration. 91% of fiscal 2026 revenue came from U.S. Government programs, and three primes, Lockheed Martin, L3Harris, and Boeing, each contributed more than 10% of consolidated sales in the same year. The mix shift out of satellite toward non-space defense programs helps, but the pipeline is still a handful of large fixed-price programs with long lead times, and one program slip can move the whole earnings line.
The timing trigger is the first quarter of fiscal 2027, the quarter in which most of the backlog is expected to be filled. That is the quarter that decides whether the digestion story holds together. The Q1 report, due in late October, is the first data point on whether gross margin recovers from the investment-heavy fiscal 2026 and whether the new Boulder, Colorado facility starts showing up in revenue. A clean print there, with the one-time charges gone and the engineering cost base absorbed, converts the thesis from plausible to demonstrated.
Frequency Electronics was founded in 1961 and builds precision time and frequency generation systems. The frequency range spans from 1 hertz to 46 gigahertz. The products go into commercial and U.S. Government satellites and into command, control, communications, intelligence, surveillance, and reconnaissance (C4ISR) and electronic warfare (EW) platforms. The company operates through two reportable segments, both anchored on Long Island, New York, with a newer satellite facility in Boulder, Colorado. The FEI-NY segment, which includes the parent and the satellite electronics and RF microwave businesses, generated 72.2% of consolidated revenue in fiscal 2026. The FEI-Zyfer segment, acquired in the early 2010s, makes precision time references for terrestrial secure communications and accounted for 34.4% of revenue before intersegment eliminations, up from 26.7% a year earlier.
The strategic frame is a deliberate reallocation from the smaller, lower-growth businesses to four larger addressable markets. Management names them explicitly: alternative position, navigation, and timing (ALT-PNT) solutions; quantum sensing, including magnetometers and Rydberg sensors; space defense and exploration; and proliferated satellite programs. The logic is that each of these markets uses the same underlying physics as atomic clocks, so the existing engineering bench can extend into them without a full technology bet. The company has made atomic clocks for decades, and a quantum sensor is a close cousin to the atomic clock in terms of the core technology, which shortens the ramp time compared to a cold-start entry.
The competitive set is a small group of niche players in atomic and optical frequency standards, a larger group of commercial oscillator and timing vendors, and the primes themselves, which can design around a supplier. The company says it is not dependent on patents and leans on technical competence and contract performance instead. In practice, the moat is a combination of space-qualified part sources, long-standing relationships with the three major satellite primes, and a track record on programs that take years to fly. The 14,000-plus operational U.S. satellites that management cites as the installed base is the anchor: every one of them carries a timing system, and the replacement and upgrade cycle, plus the growth of proliferated LEO constellations, is the structural demand story.
The government relationship is both the asset and the exposure. With 91% of sales to the U.S. Government or for U.S. Government end-use, the company is effectively a defense industrial base player at the component level. That means its revenue moves with the appropriations cycle, the procurement cycle, and the technical schedule of a small number of prime programs. The company competes for the onboard clock ensemble on GPS III options with its digital Rubidium atomic frequency standard, a program where incumbency matters more than price. The strategic risk is that the primes, which are larger and better funded, could in-house the function, and the risk is real but has not yet materialized across decades of relationships.
The product portfolio has three layers. The first and largest is satellite master timing and frequency generation, synthesis, and distribution, which includes the atomic clocks, frequency synthesizers, and distribution networks that go on GEO, MEO, and LEO spacecraft. These are long-lead, fixed-price programs where a single contract can span multiple years and carry eight-figure value. The second layer is the C4ISR and EW product line, which includes ruggedized clocks and timing references for land, sea, and airborne platforms. This is the non-space defense business, which grew 43% last year and reached about 60% of consolidated revenue. The third layer is the quantum sensing line, magnetometers and Rydberg sensors, which is early stage and has not yet contributed meaningfully to revenue.
The technical depth is the core moat. The company has held patents on quartz oscillator technology and low g-sensitivity designs, and the current patent family runs through 2026. The deeper advantage is the assembled engineering bench and the space-qualification data that takes years to accumulate. A new entrant can design a good oscillator, but qualifying it for launch across the full vibration and thermal envelope, and then proving it on an operational program, is a multi-year gate that the incumbents have already cleared. The company's R&D spend was $6.0 million in fiscal 2026, roughly a tenth of revenue. The customer-funded R&D that runs on top of that is an additional input that does not show up in the P and L as R&D expense.
The quantum sensing line is the longest-dated growth option. Rydberg sensors are compact receiving antennas that use the same physics as atomic clocks, and magnetometers are magnetic field sensors built on similar principles. The market for both is small today, but the application set, including GPS-denied positioning, electronic warfare countermeasures, and space situational awareness, is expanding. The strategic value is that the company can cross-sell into customers it already serves, and the engineering overlap means the R&D cost of entry is lower than a standalone quantum sensing company would face. The execution risk is that the product is not yet revenue-generative, and the market is still forming around what a quantum sensor is and what it costs.
The Boulder, Colorado facility, leased for 62 months starting August 2025, is the physical expression of the geographic diversification strategy. The facility is a combination office and test/assembly area, and the company expects it to be a contributor to future growth. The Long Island headquarters remains the primary production site, and the FEI-Elcom RF microwave business, which was restructured on April 30, 2026, continues under FEI-NY. The restructuring is a corporate form change, not a business exit, but it was paired with a $3.8 million inventory write-down that flowed through cost of revenues in the fourth quarter of fiscal 2026.
The most recent fiscal year was a digestion year by design. It was the year the company chose to absorb the cost of its own future. Revenue fell 9.4% to $63.2 million. The company recorded a net loss of $903,000. A year earlier the same line had shown solid net income. The income statement was hit from three directions at once: a deliberate pull-forward of satellite revenue into the prior year, the FEI-Elcom restructuring and its $3.8 million inventory write-down, and a step-up in engineering hiring and business process investment that flowed through manufacturing overhead before the corresponding revenue arrived. Gross margin compressed from 43.1% to 29.1%. Operating income swung from $11.7 million to an operating loss of $3.0 million. A tax benefit cushioned the bottom line, but it is not a recurring feature.
The segment mix tells the strategic story inside the headline numbers. FEI-NY revenue fell 14.3% to $45.7 million as the satellite pull-forward and the Elcom wind-down both landed in the same year. FEI-Zyfer revenue grew 16.5% to $21.7 million. The non-space U.S. Government defense business, which sits in both segments, grew 43.2%. It reached about 60% of consolidated revenue, up from less than half a year earlier. Satellite program revenue for Government end-use fell from 53% of total to 31%. The company is trading a higher-growth but more lumpy satellite base for a steadier, faster-growing terrestrial defense base, and the question is whether the defense business can hold the margin profile at scale.
The balance sheet was thin heading into the raise. Cash and cash equivalents at year-end stood at $1.6 million, down from $4.7 million a year earlier. Current assets of $48.0 million include $22.6 million of inventory. Contract assets add another $17.3 million. Working capital was $27.0 million. The current ratio was 2.3 to 1. The cash line itself is what the company needed to rebuild before the backlog conversion.
The late July offering added roughly a tenth of a share count. The gross raise was about a tenth of the company's annual revenue. That figure is roughly 10x the year-end cash balance. The JPMorgan revolving credit facility, signed in June, adds a committed line. Covenants sit at a total leverage ceiling and a fixed charge coverage floor. Consolidated backlog at year-end was about $111 million, compared to $70 million a year earlier. Most of it is expected to be filled next year. The nine months ended in January produced $47.8 million of revenue. Net income for the same period was $4.0 million. The run rate is already well above the prior full-year level. The company expects any partially funded contracts to become fully funded over time. The backlog excludes unfunded portions on fixed-price contracts, which means the number is a conservative floor on the pipeline. The counterargument is that a 73% fill rate in one year, on a base that includes multi-year satellite programs, still leaves room for slippage. The fixed-price accounting risk is real enough that a single cost overrun on a large program can move the whole quarter.
The central forward variable is the gross margin recovery. Fiscal 2026 gross margin of 29.1% reflects the one-time charges, the engineering hiring step-up, and the business process investment all hitting in the same period before the corresponding revenue. The company's stated position is that these investments are complete or near-complete, and the backlog ramp is what they were bought for. The first quarter of fiscal 2027 is the first clean test. If the $3.8 million inventory write-down and the one-time sick time accrual, both of which management calls non-recurring, drop out of the cost base, the gross margin line should show a material improvement even before any operating leverage kicks in. The execution risk is that the engineering cost base, once hired, does not shrink if the revenue ramp is slower than planned, and the margin recovery could be slower than the backlog growth suggests.
The second variable is the quantum sensing and ALT-PNT ramp. The company has named these markets in its annual report as the reason the Elcom wind-down was acceptable, and the engineering bench is already in place. The execution risk is that these are pre-revenue or early-revenue businesses, and the conversion from R&D investment to booked contracts takes time. The magnetometer and Rydberg sensor lines are the two products with the clearest near-term path, and the GPS-denied positioning application set is where the defense budget is moving. If the company lands a second or third design win in either category, the optionality is real. If it does not, the fiscal 2027 growth story rests entirely on the satellite and terrestrial defense base, which is still solid but does not carry the multiple-expansion narrative that the quantum line would.
The third variable is the customer mix shift. The three primes, Lockheed Martin, L3Harris, and Boeing, each contributed more than 10% of fiscal 2026 revenue, and the loss of any one would be material. The company says it is not aware of any prospect for cancellation or significant reduction, but the concentration is a structural feature, not a transient one. The non-space defense growth, at 43%, is the diversification story, and the Boulder facility is the physical expression of it. The execution risk is that the new facility, while adding capacity, also adds fixed cost, and the revenue that it is meant to support has not yet been booked.
The fourth variable is the equity position after the raise. The company raised $62.5 million at $57.50 per share, and the secondary sale by the Edenbrook funds trimmed the concentrated position that had built up over time. The board includes Jonathan Brolin, the founder of Edenbrook, and the 13G and 13D activity in the summer of 2026 reflects the fund's stake. The dilution is real, at about 10% on a pre-raise share count of 9.87 million shares, but the use of proceeds, funding the backlog ramp and the Boulder ramp, is tied to a specific growth plan rather than general corporate purposes. The counterargument is that a company with a $111 million backlog and $62.5 million of new equity is fully funded for the next two years, and the risk is not a liquidity event but an execution event.
The largest single risk is the fixed-price contract accounting risk. 95% of fiscal 2026 revenue came from fixed-price contracts, and the percentage-of-completion accounting method means that a change in estimated cost to complete can move the entire earnings line in a single quarter. The company acknowledges this in the risk factors and notes that design issues, schedule slippage, supplier issues, and labor availability can all push estimated costs up. The downside scenario is a large satellite program that comes in over budget, which would flow through the cost of revenues line and could push the company back to an operating loss for a full year. The probability is low given the company's track record, but the magnitude is high given the size of the individual programs in the backlog.
The second risk is the customer concentration. Three primes each contributed more than 10% of revenue in fiscal 2026, and the non-space defense growth, while fast, is still a minority of the total. If any one of the three primes re-competes a program, or if a large program is cancelled for technical or budget reasons, the revenue impact is material. The company's position is that the relationships are deep and the switching cost is high, which is true, but the concentration is a structural feature of the defense component supply chain and is not likely to change on a short timeframe. The downside scenario is a single program termination that removes 15% or more of the revenue base, which would be a meaningful hit to the fiscal 2027 growth plan.
The third risk is the quantum sensing timeline. The company has named this market as a core strategic priority, and the engineering investment is real, but the revenue has not yet arrived. If the quantum sensing line does not produce booked contracts within the next two fiscal years, the multiple-expansion narrative weakens, and the valuation reverts to a defense component supplier profile. The downside scenario is not a loss, but a flat revenue contribution from a market that the market has already started to price in. The company's R&D spend is not large enough to fund a standalone quantum sensing company, and the strategy depends on cross-selling into the existing defense customer base rather than building a new go-to-market motion.
The fourth risk is the government funding environment. With 91% of revenue from U.S. Government programs, the company is exposed to the appropriations cycle, the continuing resolution risk, and the procurement schedule of a small number of prime programs. A government shutdown or a significant defense budget cut would hit the company directly, and the backlog, while large, is not a contract for work performed. The company notes that its contracts are typically funded at a level less than the full contract value and require periodic incremental funding, which means the backlog is subject to the funding cycle. The downside scenario is a prolonged funding delay that pushes revenue recognition into a later quarter, which would show up as a revenue miss rather than a margin miss.
The valuation framework rests on three variables: the fiscal 2027 revenue trajectory, the gross margin recovery level, and the multiple the market is willing to assign to the combined defense and quantum sensing profile. None of the three is independently verifiable today, and that is the central feature of the setup. The market price in the mid-60s shortly after the offering implies a market capitalization near $690 million. On the share count after the raise, that is a price to book of about 10.9x on the year-end book value. The price to trailing twelve-month earnings is not meaningful given the fiscal 2026 net loss, and the forward multiple of about 40x on consensus estimates of about $1.56 per share is the relevant entry point.
The bear case assumes the gross margin recovery is slower than planned, the quantum sensing line does not contribute meaningfully to revenue in fiscal 2027, and the customer mix shift does not offset the satellite pull-forward. In that scenario, fiscal 2027 revenue lands in the low $70 million range and operating margin normalizes to the mid-teens. The net income line sits in the $4 to $5 million range. On a per share basis that is roughly $0.40. At the current price, that is a forward multiple of 140x to 180x, which is not defensible for a company without a clear multi-year growth engine. The bear case is a valuation reset to the low $40s, which is a 35% to 40% drawdown from the mid-60s.
The base case assumes the gross margin recovers to the mid-30s in fiscal 2027 as the one-time charges drop out and the engineering cost base stabilizes, the non-space defense business continues to grow at a high single-digit to low double-digit clip, and the quantum sensing line produces its first meaningful revenue contribution. In that scenario, fiscal 2027 revenue is in the mid to upper $70 million range. Operating margin is in the high teens. Net income lands in the $8 to $10 million range. On a per share basis that is roughly $0.80. At the current price, that is a forward multiple of 70x to 90x, which is rich but not irrational for a company with a $111 million backlog and a credible quantum sensing option. The base case is a price in the mid-to-high $70s, which is modest upside from the mid-60s.
The bull case assumes the quantum sensing line produces a second or third design win in fiscal 2027, the ALT-PNT market contributes meaningfully to the non-space defense growth, and the backlog fill rate holds at or above the 73% plan. In that scenario, fiscal 2027 revenue is in the low $90 million range. Operating margin is in the low 20s. Net income lands in the $14 to $16 million range. On a per share basis that is roughly $1.30. At the current price, that is a forward multiple of 45x to 55x, which is the multiple the market is already pricing in on the consensus estimate. The bull case is a price in the triple digits. That is the territory of the 52-week high. The spread between the bear and the bull is wide, and the base case is the most likely outcome, which means the current price already reflects a good deal of the upside and leaves limited cushion for a miss.
The company is in a transition, and the financials for fiscal 2026 are the transition, not the steady state. The revenue decline, the operating loss, and the gross margin compression are all explained by three specific, named, one-time events, and the backlog, which nearly doubled in the same year, is the evidence that the transition is working. The question is not whether the company has a growth story, it does, but whether the current price is already paying for it. The forward multiple of about 40x on consensus estimates is a price that assumes the gross margin recovers, the non-space defense growth holds, and the quantum sensing line starts to contribute, all in the same fiscal year. Any one of those three missing, and the multiple is too high. All three landing, and the price is reasonable.
The counterargument to the bull case is the customer concentration. Three primes each contributed more than 10% of revenue, and the defense component supply chain is a structural feature of the defense component supply chain and does not change on a short timeframe. The quantum sensing optionality is real, but it is also pre-revenue, and the market has started to price it in before the first booked contract. The valuation is not cheap, and the base case is modest upside. The bear case is a 35% to 40% drawdown, which is the risk of entering at the current price. The company is well run, the strategy is coherent, and the backlog is the real asset, but the price is already doing a lot of the work. The entry point is the 52-week low territory, not the mid-60s, and the current price is closer to the 52-week high than to the low. The investment case is a hold, not a buy, at the current price, and the next data point that matters is the Q1 fiscal 2027 report in late October, which is the first clean test of the gross margin recovery and the backlog conversion.